Dental caries lesions are a clinical manifestation of disease, preceded by microbial dysbiosis, which is poorly characterized and thought to be associated with saccharolytic taxa. Here, we assessed the associations between the oral microbiome of children and various caries risk factors such as demographics and behavioral and clinical data across early childhood and characterized over time the salivary and dental plaque microbiome of children before clinical diagnosis of caries lesions. Children ( N = 266) were examined clinically at ~1, 2.5, 4, and 6.5 y of age. The microbiome samples were collected at 1, 2.5, and 4 y. Caries groups consisted of children who remained caries free (International Caries Detection and Assessment System [ICDAS] = 0) at all time points (CFAT) ( n = 50); children diagnosed with caries (ICDAS ≥ 1) at 6.5 y (C6.5), 4 y (C4), or 2.5 y of age (C2.5); and children with early caries or advanced caries lesions at specific time points. Microbial community analyses were performed on zero-radius operational taxonomic units (zOTUs) obtained from V4 of 16S ribosomal RNA gene amplicon sequences. The oral microbiome of the children was affected by various factors, including antibiotic use, demographics, and dietary habits of the children and their caregivers. At all time points, various risk factors explained more of the variation in the dental plaque microbiome than in saliva. At 1 y, composition of saliva of the C4 group differed from that of the CFAT group, while at 2.5 y, this difference was observed only in plaque. At 4 y, multiple salivary and plaque zOTUs of genera Prevotella and Leptotrichia were significantly higher in samples of the C6.5 group than those of the CFAT group. In conclusion, up to 3 y prior to clinical caries detection, the oral microbial communities were already in a state of dysbiosis that was dominated by proteolytic taxa. Plaque discriminated dysbiotic oral ecosystems from healthy ones better than saliva.

Dentistry has entered an era of personalized/precision care in which targeting care to groups, individuals, or even tooth surfaces based on their caries risk has become a reality to address the skewed distribution of the disease. The best approach to determine a patient’s prognosis relies on the development of caries risk prediction models (CRPMs). A desirable model should be derived and validated to appropriately discriminate between patients who will develop disease from those who will not, and it should provide an accurate estimation of the patient’s absolute risk (i.e., calibration). However, evidence suggests there is a need to improve the methodological standards and increase consistency in the way CRPMs are developed and evaluated. In fact, although numerous caries risk assessment tools are available, most are not routinely used in practice or used to influence treatment decisions, and choice is not commonly based on high-quality evidence. Research will propose models that will become more complex, incorporating new factors with high prognostic value (e.g., human genetic markers, microbial biomarkers). Big data and predictive analytic methods will be part of the new approaches for the identification of promising predictors with the ability to monitor patients’ risk in real time. Eventually, the implementation of validated, accurate CRPMs will have to follow a user-centered design respecting the patient-clinician dynamic, with no disruption to the clinical workflow, and needs to operate at low cost. The resulting predictive risk estimate needs to be presented to the patient in an understandable way so that it triggers behavior change and effectively informs health care decision making, to ultimately improve caries outcomes. However, research on these later aspects is largely missing and increasingly needed in dentistry.

Objectives The aim of this study is to determine the feasibility of conducting clinical research using electronic dental record (EDR) data from U.S. solo and small-group general dental practices in the National Dental Practice-Based Research Network (network) and evaluate the data completeness and correctness before performing survival analyses of root canal treatment (RCT) and posterior composite restorations (PCR). Methods Ninety-nine network general dentistry practices that used Dentrix or EagleSoft EDR shared de-identified data of patients who received PCR and/or RCT on permanent teeth through October 31, 2015. We evaluated the data completeness and correctness, summarized practice, and patient characteristics and summarized the two treatments by tooth type and arch location. Results Eighty-two percent of practitioners were male, with a mean age of 49 and 22.4 years of clinical experience. The final dataset comprised 217,887 patients and 11,289,594 observations, with the observation period ranging from 0 to 37 years. Most patients (73%) were 18 to 64 years old; 56% were female. The data were nearly 100% complete. Eight percent of observations had incorrect data, such as incorrect tooth number or surface, primary teeth, supernumerary teeth, and tooth ranges, indicating multitooth procedures instead of PCR or RCT. Seventy-three percent of patients had dental insurance information; 27% lacked any insurance information. While gender was documented for all patients, race/ethnicity was missing in the dataset. Conclusion This study established the feasibility of using EDR data integrated from multiple distinct solo and small-group network practices for longitudinal studies to assess treatment outcomes. The results laid the groundwork for a learning health system that enables practitioners to learn about their patients' outcomes by using data from their own practice. An internet site devoted to details about the nation's network is located at http://NationalDentalPBRN.org. We are grateful to the network's regional node coordinators, who function as the “face” of the network in recruiting, training, and interacting with the network's practitioners (Midwest Region: T.S., RDH, BSDH; Western Region: Stephanie Hodge, MA; Northeast Region: Christine O'Brien, RDH; South Atlantic Region: Hanna Knopf, BA, and Deborah McEdward, RDH, BS, CCRP; South Central Region: Shermetria Massengale, MPH, CHES, and Ellen Sowell, BA; Southwest Region: Stephanie Reyes, BA, Meredith Buchberg, MPH, and Colleen Dolan, MPH), and the network's program manager (Andrea Mathews, BS, RDH), and program coordinator (Terri Jones). Our study was determined to be nonhuman subject research by the Indiana University Institutional Review Board (IRB). We also received IRB approvals or exemptions separately from the six Network regions: the Midwest, South Central, Southwest, Northeast, Western, and South Atlantic regions. The informed consent of all human subjects who participated in this investigation was obtained after the nature of the procedures had been explained fully. ^* Dr. Thankam Paul Thyvalikakath and Dr. William D. Duncan are the first authors. ^** http://www.nationaldentalpbrn.org/collaborative-group.php.

In the field of orthodontics, providing patients with accurate treatment time estimates is of utmost importance. As orthodontic practices continue to evolve and embrace new advancements, incorporating machine learning (ML) methods becomes increasingly valuable in improving orthodontic diagnosis and treatment planning. This study aimed to develop a novel ML model capable of predicting the orthodontic treatment duration based on essential pre-treatment variables. Patients who completed comprehensive orthodontic treatment at the Indiana University School of Dentistry were included in this retrospective study. Fifty-seven pre-treatment variables were collected and used to train and test nine different ML models. The performance of each model was assessed using descriptive statistics, intraclass correlation coefficients, and one-way analysis of variance tests. Random Forest, Lasso, and Elastic Net were found to be the most accurate, with a mean absolute error of 7.27 months in predicting treatment duration. Extraction decision, COVID, intermaxillary relationship, lower incisor position, and additional appliances were identified as important predictors of treatment duration. Overall, this study demonstrates the potential of ML in predicting orthodontic treatment duration using pre-treatment variables.